Evaluation of subclinical ventricular systolic dysfunction assessed using global longitudinal strain in liver cirrhosis: A systematic review, meta-analysis, and meta-regression.

Global longitudinal strain (GLS) can identify subclinical myocardial dysfunction in patients with cirrhosis. This systematic review aims to provide evidence of a possible difference in GLS values between patients with cirrhosis and patients without cirrhosis. Studies from inception to August 11, 2021, were screened and included based on the inclusion criteria. The Newcastle Ottawa Scale was used to assess the quality of nonrandomized studies. Meta-analyses were conducted with subsequent sensitivity and subgroup analyses according to age, sex, cirrhosis etiology, and severity. Publication bias was evaluated using Begg's funnel plot, Egger's test, and rank correlation test with subsequent trim-and-fill analysis. The systematic database search yielded 20 eligible studies. Random effect showed a significant reduction of left ventricular (LV) GLS (MD:-1.43;95%; 95%CI,-2.79 to -0.07; p = 0.04; I2 = 95% p<0.00001) and right ventricular (RV) GLS (MD:-1.95; 95%CI,-3.86 to -0.05, p = 0.04; I2 = 90%, p<0.00001) in the group with cirrhosis. A sensitivity test on subgroup analysis based on the study design showed a -1.78% lower LV-GLS in the group with cirrhosis (I2 = 70%, p = 0.0003). Meta-regression analysis showed that the severity of cirrhosis was significantly related to GLS reduction. This research received no specific grants from any funding agency in the public, commercial, or not-for-profit sectors. The study protocol was registered at PROSPERO (CRD42020201630). We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement guidelines.

Introduction

Cirrhotic cardiomyopathy (CCM) is defined by the World Congress of Gastroenterology [1], revised in 2019 according to criteria from the Cirrhotic Cardiomyopathy Consortium, as a left ventricular ejection fraction (LVEF) less than or equal to 50% and/or absolute global longitudinal strain (GLS) less than 18% in patients with cirrhosis [2]. A hyperdynamic state, characterized by increased cardiac output and decreased peripheral vascular resistance, occurs in patients with liver cirrhosis [3]. This hyperdynamic condition can progress to abnormal contractile response to stress and/or altered diastolic relaxation, with electrophysiological abnormalities in the absence of known cardiac disease [4-6]. Myocardial stiffness and subsequent hypertrophy due to sodium and fluid retention can also lead to diastolic dysfunction [7,8]. CCM is often under-recognized, although it is acknowledged that early recognition is important in determining prognosis, especially for patients who may require future procedures that could potentially stress the heart, such as shunt implantation and liver transplantation [9].

Many modalities are used to measure myocardial strain, or regional deformation of the myocardium, and evaluate myocardial performance [5,10,11]. Myocardial strain measurements can be a better quantitative approach compared with the conventional ejection fraction (EF) measurement, which is highly dependent on geometrical assumptions and endocardial border definition [11,12]. EF evaluation is limited by reader experience and does not consider that a hyperkinetic segment may compensate for a hypokinetic one, leading to a false “normal” result [12]. These limitations decrease the sensitivity of EF to diagnose myocardial dysfunction at an early stage [11]. Speckle tracking echocardiography (STE) has become the preferred modality to measure myocardial based on the change in myocardial length in one region [11] Myocardial strain can be classified into circumferential, longitudinal, radial, and transverse strains; these are better at evaluating contractile function than left ventricular ejection fraction (LVEF) alone, which primarily reflects radial function [2]. Global longitudinal strain (GLS) analysis using STE has demonstrated the ability to identify subclinical myocardial dysfunction in various diseases [13]. The use of GLS to identify myocardial contractile dysfunction in patients with preserved LVEF is important because longitudinal contractile function is often impaired before radial contractile function [2]. Other modalities, such as cardiovascular magnetic resonance (CMR), have also been utilized as a standard reference for evaluating cardiac morphology, volume, and myocardial strain [10]. According to the latest CCM consortium [2], data on strain imaging to detect CCM in patients with normal LVEF are limited and conflicting, with three studies showing normal longitudinal strain [10,14,15] and one multi-center study showing diminished longitudinal strain in one of two cohorts of patients with cirrhosis [16]. Thus, our systematic review aimed to evaluate whether GLS values differ between patients with cirrhosis and patients without cirrhosis.

Materials and methods

The study protocol was registered at the International Prospective Register of Systematic Reviews (PROSPERO) with registration number CRD42020201630 [17]. This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 guidelines for reporting [18]. Furthermore, we complied with the guidelines for conducting systematic reviews and meta-analyses of observational studies of etiology (COSMOS-E) [19]. We used a pre-determined search strategy to conduct a structured search of the literature to identify studies on the outcome of GLS in patients with cirrhosis versus patients without cirrhosis. The search was conducted in the Cochrane Library, EBSCOhost, Open Grey, PMC, ProQuest, PubMed, and ScienceDirect databases from inception to August 11, 2021; additionaly, a manual search was performed to retrieve relevant studies. Using MeSH terms and [All Field], we complemented the search strategy using the following keywords: ventricular function, ventricular dysfunction, myocardial, ultrasonography, echocardiography, cardiac magnetic resonance, speckle tracking, longitudinal strain, tissue-Doppler, liver cirrhosis, and end-stage liver disease.

The results of the search strategy were exported to Endnote X9, duplicates were removed, and the remaining articles were reviewed based on the title and abstract. Studies were included based on the following criteria: (1) observational studies with participants aged > 18 years; (2) article written in English; and (3) availability of GLS data from patients without cirrhosis versus patients with cirrhosis estimated using mean or median. Studies were excluded if: (1) study participants aged > 80 years; (2) participants with reduced/mid-range EF heart failure as defined by the European Society of Cardiology 2016 guideliens [20]; (3) no control groups available for complete data extraction; (4) data for ventricular longitudinal strain not available; and (5) full articles not retrieved. After selection of all studies fulfilling the inclusion criteria, the following data were extracted: first author, publication year, country of origin, sample size, age of participants, and GLS values. For duplicate data extraction, two authors performed data extraction to reduce the possibility of a single person’s bias. The study authors were contacted via email to request access to missing data. The mean and median estimations of GLS were assessed to elucidate possible differences between patients with and without cirrhosis. The mean difference and 95% confidence interval (CI) were used to determine the difference between the compared groups. Thus, all studies reporting median estimations were approximated into mean estimations using the method proposed by Wan et al., which performs very well for both normal and skewed data [21].

To detect statistical heterogeneity, we used Cochrane’s Q test (chi-squared test) and Higgins I2 statistics. Heterogeneity was considered to be present if P < 0.10 or I2 > 75% [22,23]. A forest plot was generated to evaluate heterogeneity. If heterogeneity was present, we performed a random-effects model (REM) using the DerSimonian-Laird method [24].

Publication bias analyses were performed using JASP version 0.16.1 (JASP Team, Amsterdam) [25]. Begg’s funnel plot was generated to assess publication bias when the number of included studies was at least 10 and heterogeneity was not statistically significant [22,26]. This was further confirmed by Egger’s test and Begg and Mazumdar’s rank correlation test [27,28]. Correction of publication bias was based on Duval and Tweedie’s trim-and-fill method if the heterogeneity was less than 75% [26,29]. The overall fail-safe number of publications to assure a borderline significant effect size was calculated according to Rosenthal et al [30]. When the fail-safe number was relatively large compared to the number of included studies, higher confidence was assured regarding the stability of the results. A fail-safe number is considered robust if it is five times higher than the number of studies included, plus 10 [30].

We performed a random-effects (method of moments) meta-regression using Comprehensive Meta-Analysis version 3.3.070. software (Biostat Inc., Englewood, NJ, USA) [31]. Meta-regression was conducted to investigate the true causes of heterogeneity that explained the high value of the I2 statistic [32]. Sources of potential variability were based on covariates of study design, proportion of male subjects, mean age of study sample, mean model for end-stage liver disease (MELD) score, proportion of decompensated cirrhosis (Child-Pugh class B and C), proportion of patients with alcoholic-etiology cirrhosis, proportion of patients with viral-etiology cirrhosis, baseline LVEF in the group with cirrhosis, and methodological quality of the study.

For sensitivity analysis, we deleted one study at a time to determine the effect and stability of one study on the pooled mean difference [33]. Subgroup analysis was performed when at least 10 studies were included [33]. Subgroup analysis was performed based on the study design, despite no strong evidence of statistical heterogeneity [34]. Subgroup analysis was also performed based on age, sex, cirrhosis etiology, and severity (MELD or Child-Pugh classification) to detect clinical heterogeneity. These variables were chosen because previous studies support the potential effects of age [2,35], sex [2], cirrhosis etiology [36], and severity [2,37-40] on GLS results. For cirrhosis severity, if provided by at least 10 studies [33], subgroup analysis was performed by dividing patients with cirrhosis into compensated (Child-Pugh class A) and decompensated (Child-Pugh class B and C) groups as classified in a previous study [36]. Furthermore, sensitivity analysis was also performed within each subgroup analysis to reduce heterogeneity to less than 75% by omitting the study with the lowest methodological quality while also having the largest heterogeneity contribution. Meta-analysis was performed using Review Manager 5.3.5 software (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration) [41].

The methodological quality of observational studies was in accordance with the Newcastle Ottawa Scale (NOS) [42]. The NOS assesses participant selection, comparability, and outcome reporting using eight subscale items [42]. For cross-sectional studies, an adapted version of NOS was used, similar to previous studies [43,44]. The sum of the subscale item scores, with a maximum of 10, was used to provide an overall assessment of evidence quality for each study. For case-control and cohort studies, a maximum score of nine from the sum of the subscale items was used [42]. All risk of bias analyses were performed by two authors, and disagreements were resolved by a third author.

Results

The search strategy identified 5347 studies from the database searches (S1 Table), and the manual search identified seven additional studies. The results were imported into Endnote X9, and duplicates were removed, leaving 5068 articles for review. The article abstracts were reviewed for relevance based on inclusion and exclusion criteria. After screening, 26 studies from the databases and seven studies from manual searching methods were retained for full review. Of these, 20 were retained for the analysis (Fig 1). The other 13 studies were excluded for the following reasons: full-text failed to be retrieved in two studies, comparison with patients without cirrhosis was unavailable in nine studies, ventricular longitudianl strain was not analyzed in one study, and population with EF < 50% in one study.

Fig 1

PRISMA 2020 flow diagram of the identification and selection of studies included in the analysis.

This systematic review identified six cross-sectional [11,15,45-48], 6 case-control [10,49-53], and eight prospective cohort studies[14,36,54-59] gathered from 12 countries and consisting of 1738 participants (Table 1). Our search strategy did not identify any randomized controlled trials. We found one cross-sectional [15], one case-control [53], and three prospective cohort studies [14,55,59] providing results for GLS in both ventricles. There were four cross-sectional [11,45,46,48], five case-control [10,49-52], five prospective cohort studies [36,54,56-58] providing results only for the left ventricle. Meanwhile, one cross-sectional study [47] providing result only for the right ventricle. There were two case-control studies [10,52] in which evaluations were performed using CMR and others using 2D-STE. Five cross-sectional [11,15,45,47,48], three case-control [50,51,53], and five prospective cohort studies [14,36,54,57-59] used the EchoPAC system; one cross-sectional [46], one case-control [49], and one prospective cohort study [55] used Velocity Vector Imaging and one prospective cohort study [56] used the QLAB system for 2D-STE evaluation. The etiology of cirrhosis varied; the most common etiologies were hepatitis B and C viruses [10,14,15,36,45,46,48,49,52-54,56-59]. Three studies did not specify the etiology [11,51,55]. Comorbidities outside cardiac pathology among the cirrhotic group included hypertension and diabetes mellitus, which were reported in one cross-sectional [46] and two prospective cohort [14,57] studies. Other comorbidities such as dyslipidemia and pulmonary artery hypertension were reported in one cross-sectional [46] and one case-control study [51], respectively.

Table 1

Summary of included studies with left and right ventricular global longitudinal strain results.

Author (Year)	Subjects (LC [%male] vs non-LC [%male])	Baseline Ages (LC vs non-LC, years old) [Median, Q1-Q3 / Mean ± SD])	Cirrhosis Etiology n (%)	Cirrhosis Prognostic Score and Severity	Cirrhosis Group Comorbidity n (%)	Left Ventricular Longitudinal Strain (LC vs non-LC, %) (Median, Q1-Q3 / Mean ± SD)	Right Ventricular Longitudinal Strain (LC vs non-LC, %) [Median, Q1-Q3 / Mean ± SD]	Other Parameter (TAPSE [mm], RVFAC [%]) (LC vs non-LC)
CROSS-SECTIONAL
Hammami R (2017) [45]	80 [52.5] vs 80 [N.R]	55 ± 14 vs 51 ± 12 (p>0.05)	Viral (Hepatitis B and C): 42 (52.6%)Cryptogenic: 21 (26.1)Other causes: 17 (21.25)	Mean MELD Score: 14.2 ± 4.98Mean CP Score: N.RCP classification n (%):A: 24 (30)B: 36 (45)C: 20 (25)	Left ventricular hypertrophy: 39 (48.75)	Apical 4,2,3-chamber view: -19.8 ± 2.8 vs -22.01 ± 2.6 (p<0.001)	N.R	N.R
Rimbaş RC (2017) [15]	46 (65.2) vs 46 (72.5)	57 ± 9 vs 55 ± 10 (p>0.05)	Alcoholic = 52%Viral = 41%Primary biliary cirrhosis = 4%Cryptogenic = 2%	Mean of MELD Score:13 ± 5Mean of CP Score: 7 ± 2CP classification (n,%):A: 23 (50)B: 16 (35)C: 7 (15)	Diastolic dysfunction: 22 (47.8)	Apical 4,2,3-chamber and short axis view at level of papillary muscle:-20.9 ± 3 vs -20.7 ± 2.8 (p>0.05)	Apical 4-chamber view:-24 ± 5 vs -23 ± 4 (p>0.05)	TAPSE: 26 ± 5 vs 25 ± 3 (p>0.05)RVFAC: 43 ± 12 vs 44 ± 8 (p>0.05)
Novo G (2018) [46]	39 [41.02] vs 39 [43.58]	60 (48‐70) vs 59 (48‐67) (p>0.05)	Hepatitis C: 39 (100)	Median of MELD score: 7 (6–8)Median of CP score: N.RCP classification (n,%):A: 39 (100)	Hypertension: 17 (43.58)Diabete mellitus: 11 (28.20)Dyslipidemia: 2 (5.13)	Apical 4,2,3-chamber view:Median: -18.1 (‐20.5 to -16.3) vs -21.2 (‐22.3 to -20.4) (p = 0.001)Mean estimation (Wan’s method): -18.3 ± 3.23 vs -21.3 ± 1.46 (p<0.001)	N.R	TAPSE: 22 (21‐24) vs 23 (20‐25) (p = 0.702)Mean Estimation (Wan’s method): 22.33 ± 2.31 vs 22.67 ± 3.85 (p>0.05)
Zamirian M (2019) [11]	20 [50] vs 10 [80]	42.2 ± 4.7 vs 41.6 ± 4.7 (p>0.05)	N.R	Mean of MELD score: N.RMean of CP score: N.RCP classification (n,%):C: 20 (100)	Diastolic dysfunction: 4 (20)	Apical 4,2,3-chamber view:-22.6 ± 2.4 vs -19.2 ± 1.9 (p = 0.001)	N.R	N.R
Zhang K (2019) [47]	67 [44] vs 36 [49]	53 ± 12 vs 47 ± 13 (p>0.05)	Alcoholic: 67 (100)	Mean of MELD score: 13.5 ± 8.9Mean of CP score: N.RCP classification (n,%): N.R	N.R	N.R	Apical 4- chamber view:-19.8 ± 4.2 vs -21.8 ± 1.7 (p = 0.005)	TAPSE: 25 ± 5 vs 25 ± 4 (p = 0.809)RVFAC: 48 ± 10 vs 47 ± 8 (p = 0.483)
von Köckritz F (2021) [48]	80 [58.8] vs 30 [46.7]	52.47 ± 10.24 vs 48.57 ± 12.93 (p = 0.145)	Alcoholic: 31.25%Hepatitis C: 12.5%Autoimmune: 10%NASH: 10%PSC: 8.75%,Idiopathic: 8.75%Other cause (cystic liver, Wilson’s disease, bile duct carcinoma, and Caroli syndrome): 18.75%	Mean of MELD score: 17 ± 6.65Mean of CP score: N.RCP classification (n,%):C: 80 (100)	Diastolic dysfunction: 14 (17.5)	Apical 4,2,3-chamber view:-21.39 ± 4.06 vs -18.73 ± 2.95 (p<0.001)	N.R	N.R
CASE-CONTROL
Sampaio F (2013) [49]	109 [78.9] vs 18 [17.2]	54 (48–64) vs 51 (49–58) (p>0.05)	Alcoholic: 73 (67)Viral: 27 (24.8)Other: 9 (8.2)	Median of MELD score: 14 (10–18)Median of CP Score: N.RCP classification (n,%):A: 37 (33.9)B: 27 (24.8)C: 45 (41.3)	Diastolic dysfunctoin: 44 (40.3)	Apical 4,2-chamber view:-19.99% (-21.88 to -18.71) vs -22.02% (-23.10 to -21.18) (p = 0.003)Mean estimation (Wan’s method): -20.19 ± 2.38 vs -22.10 ± 1.54 (p<0.05)	N.R	TAPSE: 25.4 mm (22.0–28.2) vs 23.1 (21.5–26.2) (p = 0.11)Mean Estimation (Wan’s method): 25.2 ± 4.66 vs 23.6 ± 3.78 (p>0.05)
Al-Hwary S (2015) [50]	20 [N.R] vs 40 [N.R]	46.45 ± 6.29 vs 43.25 ± 5.11 (p>0.05)	N.R	Mean of MELD score: N.RMean of CP score: N.RCP classification (n,%): All patients are stable cirrhotic patients	Hypertension	Apical 4,2,3-chamber view:-19.98 ± 7.65 vs -29.50 ± 5.92 (p<0.05)	N.R	N.R
Sampaio F (2015) [10]	36 [83.3] vs 8 [62.5]	54 (48–61) vs 52 (45–54) (p>0.05)	Alcoholic: 21 (58.3)Viral: 10 (27.8)Other causes: 5 (13.9)	Median of MELD score: 9 (7–11)Median of CP Score: 5 (5–7)CP classification (n,%)A: 27 (75)B: 8 (17.8)C: 1 (7.2)	N.R	Apical 4,2,3-chamber view:Median: −18.9 (-16.0 to -20.5) vs -19.0 (-16.1 to -20.6) (p = 0.96)Mean estimation (Wan’s method): -18.47 ± 3.47 vs -18.57 ± 4.02 (p = 0.94)	N.R	N.R
Anish PG (2019) [51]	55 [83.63] vs 30 [83.33]	46.38 vs 45.56 (p>0.05)	N.R	Mean of MELD score: 12MELD score > 12 (n,%): 22 (40)MELD score < 12 (n,%): 33 (60)Mean of CP score: N.RCP classification (n,%): N.R	Pulmonary artery hypertension: 18 (32.7)Left ventricular hypertrophy: 26 (47.3)	Apical 4,2,3-chamber view:-19.52 ± 2.41 vs -23.66 ± 2.31 (p<0.0001)	N.R	N.R
Isaak A (2020) [52]	42 55 vs 18 [72]	57 ± 11 vs 54 ± 19 (p>0.05)	Alcoholic: 24 (57)Viral hepatitis: 4 (12)Autoimmune hepatitis: 3 (7)NASH: 3 (7)Hemochromatosis: 1 (2)Congenital anomaly: 1 (2)Cryptogenic: 5 (12)	Mean of MELD score:CP A: 10 ± 2CP B: 12 ± 5CP C: 15 ± 5Mean of CP score: N.RCP classification (n,%)A: 11 (26)B: 20 (48)C: 11 (26)	N.R	Apical 4,2-chamber and parasternal short axis views:-18.5 ± 4.0 ± -22.5 ± 3.6 (p<0.001)	N.R	N.R
Koç DÖ (2020) [53]	50 [62] vs 33 [51.5]	57 ± 13 vs 55 ± 12 (p>0.05)	Viral Hepatitis: 30 (60)NAFLD: 12 (24)Alcoholic: 4 (8)Other cause: 4 (8)	Mean of MELD score:15.84 ± 7.92MELD score > 15 (n,%): 25 (50)MELD score < 15 (n,%): 25 (50)Mean of CP score: N.RCP classification (n,%)A: 19 (38)B: 14 (28)C: 17 (34)	N.R	Apical 4,2-chamber view:-19.42 ± 2.83 vs -19.49 ± 2.33 (p>0.05)	Basal, middle, apical segment and ventricular septum view:17.05 ± 3.49 vs 22.61 ± 0.93 (p = 0.001)	N.R
PROSPECTIVE COHORT
Altekin RE (2014) [54]	38 [63.2] vs 37 [54.1]	48.3 ± 12.4 vs 45.4 ± 8.6 (p>0.05)	Viral (Hepatitis B and C): 23 (60.5)Cryptogenic: 10 (26.3)Biliary: 5 (13.1)	Mean of MELD score: 11.76 ± 4.92Mean of CP score: N.RCP classification (n,%):A: 23 (60.5)B: 12 (31.6)C: 3 (7.9)	N.R	Apical 4,2,3-chamber and parasternal short axis view:-20.57 ± 2.1 vs -28.74 ± 3.11 (p<0.001)	N.R	N.R
Huang CH (2019) [36]	80 [80] vs 29 [65.5]	48.5 (45.0–59.0) vs 49.0 (43.0–52.5) (p>0.05)	Alcoholic: 28 (25.7)Hepatitis B: 22 (20.2)Hepatitis C: 30 (27.5)	Mean of MELD score:Liver cirrhosis with CCM (n = 22): 15.9 ± 8.3Liver cirrhosis without CCM (n = 57): 15.3 ± 7.9Mean of CP score: N.RCP classification (n,%):A: 31 (38.75)B/C: 49 (61.25)	Diastolic dysfunction: 27 (34.2)	Apical 4,2,3-chamber view:Median: -21.5 (-22.4 to -20.4) vs -20.2 (-23.0 to -19.1) (p = 0.108)Mean estimation (Wan’s method): -21.43 ± 1.51 vs -20.77 ± 3.04 (p = 0.136)	N.R	N.R
İnci SD (2019) [55]	40 [70] vs 26 [61.54]	46.2 ± 10.1 vs 42.2 ± 8.6 (p>0.05)	N.R	Mean of MELD score: N.RMean of CP score: N.RCP classification (n,%):C: 40 (100)	N.R	Apical 4-chamber view: -16.0 ± 3.2 vs -17.6 ± 2.2 (p = 0.003)Apical 2-chamber: -16.2 ± 3.3 vs -18.7 ± 2.1 (p = 0.002)	Apical 4- chamber view:-19.2 ± 3.5 vs -21.5 ± 3.6 (p = 0.003)	N.R
Özdemir E (2019) [56]	40 [33] vs 40 [33]	42.8 ± 8.8 vs 42.5 ± 11.4 (p>0.05)	Hepatitis B = 40 (100)	Mean of MELD score: N.RMean of CP score: N.RCP classification (n,%): N.R	N.R	Apical 4,2,3-chamber and parasternal short axis view:-19.9 ± 3.4 vs -22.8 ± 1.9 (p<0.001)	N.R	N.R
Kim HM (2020) [57]	33 [75.8] vs 17 [55]	56.3 ± 9.9 vs 65.0 ± 14.8 (p>0.05)	Viral (Hepatitis B and C): 20 (60.6%)Alcoholic: 9 (27.3)Autoimmune hepatitis: 2 (6.1)Cryptogenic: 2 (6.1)	Mean of MELD score: 18.8 ± 7.4Mean of CP score: 9.8 ± 2.4CP classification (n,%)A/B: 10 (30.3)C: 23 (69.7)	Hypertension: 8 (24.2)Diabetes mellitus: 9 (27.3)	Apical 4,2,3-chamber view:-24.2 ± 2.7 vs -18.6 ± 2.2 (p<0.001)	N.R	N.R
Chen Y (2016) [14]	103 [74.8] vs 48 [66.7]103 cirrhotic patients were classified into:Undergoing LTx: 41Without LTx: 26Refusing Echo follow-up: 14Died during study period: 22	54.9 ± 7.3 vs 53.5 ± 7.9 (p>0.05)	Undergoing LTx (n = 41)Alcoholic: 6 (14.6)Viral: 30 (73.2)Others: 5(12.2%)Without LTx (n = 26)Alcohol: 3 (11.5)Viral: 17 (65.4)Others: 6(23.1)Refusing Echo follow-up: N.RDied during study period: N.R	Undergoing LTx (n = 41)Mean of MELD score: 21.3 ± 8.9Mean of CP Score: N.RCP classification (n,%):A: 7 (17.1)B: 11 (26.8)C 23 (56.1)Without LTx (n = 26)Mean of MELD score: 12.2 ± 5.6Mean of CP Score: N.RCP classification (n,%):A: 12 (46.2)B: 10 (38.5)C: 4 (15.3)Refusing Echo follow-up: N.RDied during study period: N.R	Hypertension: 26 (25.2%)Diabetes mellitus: 21 (20.4)	Apical 4,2,3-chamber view:-18.6 ± 2.6 vs -20.1 ± 2.8 (p<0.01)	Apical 4-chamber view:-21.2 ± 4.4 vs -23.0 ± 2.6 (p<0.01)	TAPSE: 23 ± 4 vs 23 ± 2 (p = 0.77)RVFAC: 53 ± 8 vs 55 ± 6 (p = 0.06)
Hassan AA (2019) [58]	45 [42] vs 30 [53]	47.13 ± 9.2 vs 46.8 ± 8.9 (p>0.05)	Hepatitis C: 45 (100)	Mean of MELD score: N.RMean of CP score: N.RCP classification (n,%):A: 15 (33)B: 15 (33)C: 15 (33)	N.R	Apical 4,2,3-chamber view:-19.5 ± 2.7 vs -20.7 ± 4.3 (p = 0.04)	N.R	N.R
Ibrahim MG (2020) [59]	50 [42] vs 50 [38]	52 ± 12.04 vs 46.76 ± 12.1 (p>0.05)	Hepatitis C: 50 (100)	Mean of MELD score: N.RMean of CP score: N.RCP classification (n,%):A: 38 (76)B: 12 (24)	Hypertension: 28 (56)Diabetes mellitus: 14 (28)Chronic Hepatitis C infection: 50 (100)	Apical 4,2,3-chamber view:Median: -20 (-26 to -16.5) vs -20 (-28 to -17)Mean estimation (Wan’s method): -20.83 ± 7.25 vs -21.67 ± 8.4 (p = 0.59)	Apical 4-chamber view:Median: -22 (-30 to -17) vs -22 (-30 to -17)Mean estimation (Wan’s method): -23 ± 9.92 vs -23 ± 9.92 (p = 1.00)	TAPSE: 24.56 ± 3.08 vs 24.06 ± 2.65 (p>0.05)RVFAC: 45.72 ± 4.88 vs 45.64 ± 4.89 (p>0.05)

ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CCB, calcium channel blocker; LC, liver cirrhosis; LTx, liver transplantation; MELD, model for end-stage liver disease; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; N.R, not reported; PSC, primary sclerosing cholangitis; RVFAC, right ventricular fractional area change; TAPSE, tricuspid annular plane systolic excursion.

All five prospective cohorts had different durations of follow-up, including 7 [54], 8[56], 12 [36], 14 [58], 19 [59], 22 [57], and 32 months [14]. One study [55] did not report the follow-up duration. Altekin et al. [54] included patients with cirrhosis between the ages of 20 and 65 years, with 60.5% in the compensated group and 39.5% in the decompensated group. The most common etiology was viral hepatitis B, with a mean MELD score of 11.76. Huang et al. [36] included patients with cirrhosis between the ages of 35 and 65 years, with 38.75% in the compensated group and 61.25% in the decompensated group. The most common etiology was viral hepatitis B and C, with mean MELD score of 15.47. Inci et al. [55] included patients with cirrhosis between the ages of 30 and 60 years, with all subjects in the decompensated group. In this study, the etiology and mean MELD scores were not reported. Özdemir et al. [56] included patients with cirrhosis aged 30–60 years. The proportion of patients with compensated or decompensated cirrhosis has not yet been reported. All cirrhosis etiologies were due to hepatitis B, and the mean MELD score was not reported. Kim et al. [57] included patients with cirrhosis between the ages of 40 and 70 years, with 30.3% in Child-Pugh A and B, and 69.7% in Child-Pugh C. The most common etiology was hepatitis B, with a mean MELD score of 18.8. Chen et al. [14] included patients with cirrhosis between the ages of 40 and 70 years, with 71.64% in the decompensated group. The most common etiology was viral infection, with a mean MELD score of 17.77. Hassan et al. [58] included patients with cirrhosis between the ages of 20 and 65 years, with 66.66% in the decompensated group. All cirrhosis etiologies were hepatitis C. Mean of MELD score was not reported. Ibrahim et al. [59] included patients with cirrhosis between the ages of 30 and 70 years, with 24% in the decompensated group. All cirrhosis etiologies were hepatitis C. The mean MELD score was not reported.

For left ventricular assessment, there were two cross-sectional [45,46], four case-control [49-52], and four prospective cohort studies [14,54-56] showing a significantly lower absolute GLS in patients with cirrhosis than in patients without cirrhosis. Meanwhile, two cross-sectional studies [11,48] and one prospective cohort study [57] reported a higher absolute GLS in patients with cirrhosis than in patients without cirrhosis. We found one cross-sectional study [15], one case-control study [10], and three prospective cohort studies [36,58,59] that reported neutral results. For right ventricular assessment, there were one cross-sectional [47], one case-control [53], and two prospective cohort [14,55] studies showing a significantly lower absolute GLS in patients with cirrhosis than in patients without cirrhosis. Meanwhile, one cross-sectional [15] and one prospective cohort [59] study reported neutral results. Most studies reported an EF > 55% in groups with and without cirrhosis. Considering the standard deviation, six studies [15,45,53,56,58,59] included patients with a borderline EF (50–55%).

Methodological quality for cross-sectional studies

Methodological quality scores for cross-sectional studies were 8/10 and 9/10 (S3 Table). None of the studies provided information about sample size calculations. The study by Rimbaş et al. [15] excluded 29 of 75 patients with cirrhosis but did not describe the characteristics of these non-respondents. The study by Zhang et al. [47] included only alcoholic cirrhosis; meanwhile, the study by Novo et al. [46] included only patients with hepatitis C etiology and Child-Pugh class A severity cirrhosis. The studies by Zamirian et al. [11]. and von Köckritz et al. [48] included only patients with Child-Pugh class C severity.

Methodological quality for case-control studies

Methodological quality scores for the case-control studies were 7/9 and 8/9 (S4 Table). None of these case-control studies provided a clear description of the non-response rate. Sampaio et al. [49] and Isaak et al. [52] reported the number of excluded participants in the group with cirrhosis, but not in the control group, and a large discrepancy in the population was evident between the case and control group. In addition to a lack of reports concerning the non-response rate, Al-Hwary et al. [50] did not provide any information regarding the validation of the cirrhosis diagnosis.

Methodological quality for cohort studies

Methodological quality scores ranged from 6/9 to 9/9 (S5 Table). The study by Altekin et al. [54] scored 6/9 because it excluded alcoholic cirrhosis and thus did not cover the entire population of cirrhotic liver disease. In addition, this study had no description of the non-exposed cohort derivation and a shorter follow-up period (less than 6 months) compared with other studies (at least 1 year). Meanwhile, the study by Kim et al. [57] scored 6/9 because there was no description of the source of the non-exposed cohort and no adjustment for confounders, stratifications, or matching to improve comparability between the group with cirrhosis and control. The study by İnci et al. [55] only included patients with Child–Pugh C class severity cirrhosis and did not report the length of follow-up. Özdemir et al. [56] only included patients with hepatitis B viral infection etiology and did not report the source of non-exposed group derivation. Chen et al. [14] reported a different source of population derivation between the exposed and non-exposed cohorts. Hassan et al. [58] evaluated only the group with hepatitis C viral infection cirrhosis etiology and did not report the source of non-exposed group derivation. The study by Ibrahim et al. [59] included only patients with hepatitis C viral etiology cirrhosis.

Left ventricular global longitudinal strain in patients with cirrhosis versus patients without cirrhosis

We evaluated 19 studies [10,11,14,15,36,45,46,48-53,55-59] reporting LV-GLS from patients with and without cirrhosis. The pooled analysis of LV-GLS, evaluated using 2D-STE in17 studies and CMR in two studies, revealed a significantly lower absolute LV-GLS based on the mean difference in patients with cirrhosis (MD:-1.43; 95%CI,-2.79 to -0.07; p = 0.04) (Fig 2). REM was considered to be due to the significant statistical heterogeneity between studies (I2 = 95%,p<0.00001).

Fig 2

Mean difference of left ventricular global longitudinal strain in patiens with cirrhosis versus patients without cirrhosis evaluated using the random effect model.

SD , standard deviation; IV ,  inverse variance; CI ,  confidence interval; df ,  degrees of freedom; Chi2 , chi-squared statistic; p ,  p-value; I2 ,  I-squared heterogeneity statistic; Z , Z statistic.

We conducted sensitivity analysis to evaluate the stability of our findings. The pooled results changed several times after each omission of one study from the meta-analysis (S6 Table). The omission of 10 studies [14,45,46,49-52,54-56] individually revealed insignificance in the pooled meta-analysis. Meanwhile, omission of the other nine studies [10,11,15,36,48,53,57-59] resulted in a significantly lower absolute LV-GLS in favor of the group with cirrhosis. We found that the omission of the study by Altekin et al. [54] resulted in the largest Q and Tau2 change. The results showed no significant absolute LV-GLS difference (MD:-1.02; 95%CI,-2.21 to 0.18; p = 0.09; I2 = 93%, p<0.00001) (S1 Fig). Further sensitivity analysis was performed, and the omission of the study by Kim et al. [57] showed the largest change in Q and Tau2 (S7 Table). Despite the persistent heterogeneity, the final result successfully provided stability in the next sensitivity analysis (S8 Table), resulting in a significantly lower absolute LV-GLS favoring the group with cirrhosis (MD:-1.39; 95%CI,-2.40 to -0.37; p = 0.008; I2 = 90%, p<0.00001) (S2 Fig). To achieve a heterogeneity of less than 75%, seven studies [11,36,48,50,51,54,57] were omitted, and the pooled meta-analysis from the resulting 12 studies showed a significantly lower absolute LV-GLS favoring the group with cirrhosis (MD:-1.66; 95%CI,-2.33 to -1.00; p <0.00001; I2 = 69%, p = 0.0002) (Fig 4). Moreover, to evaluate whether cardiovascular comorbidities could contribute to a decreased LV-GLS, we excluded studies that included patients with hypertension, diabetes mellitus, and left ventricular hypertrophy. Further exclusion of three studies [14,46,59] that included patients with hypertension and diabetes mellitus showed similar results (MD:-1.55; 95%CI,-2.37 to -0.73, p = 0.0002; I2 = 73%, p = 0.0003), while exclusion of the study by Hammami et al. [45] that included patients with left ventricular hypertrophy showed simillar results (MD:-1.59; 95%CI,-2.34 to -0.84, p<0.00001; I2 = 70%, p = 0.0002). Other comorbidities, such as coronary artery disease and valvular heart disease, were excluded in all studies; therefore, these variables did not influence the current results. Additionally, the exclusion of two studies [15,49] which included patients with diastolic dysfunction revealed a similar result, indicating a significant reduction in LV-GLS in patients with cirrhosis (MD:-1.85; 95%CI,-2.57 to -1.33, p<0.00001; I2 = 63%, p = 0.003). In order to evaluate whether ongoing medical therapy (diuretics, beta-blockers, angiotensin-converting enzyme inhibitor [ACE]-I/angiotensin receptor blocker [ARB], calcium channel blocker [CCB]) could influence LV-GLS, we excluded three studies [10,14,45] that reported these medications in patients with cirrhosis, and the result showed that there was still a significant reduction of LV-GLS in patients with cirrhosis (MD:-1.69; 95%CI,-2.59 to -0.79, p = 0.0002; I2 = 76%, p<0.0001). The effect of diuretic treatment seen in the study by Sampaio et al. [10] was excluded from the analysis, resulting in a similar result (MD:-1.72; 95%CI,-2.40 to -1.04, p<0.00001; I2 = 71%, p = 0.0002). Excluding the study by Hammami et al. [45] to evaluate the influence of beta-blocker treatment on the results also showed similar findings (MD:-1.59; 95%CI,-2.34 to -0.84, p<0.0001; I2 = 70%, p = 0.0002).

Fig 4

Mean difference of left ventricular global longitudinal strain in patients with cirrhosis versus patients without cirrhosis after consecutive omission of studies by Altekin, et al.; Kim, et al; Zamirian, et al; von Köckritz, et al.; Anish, et al.; Al-Hwary, et al.; and Huang, et al.

SD , standard deviation; IV , inverse variance; CI ,  confidence interval; df ,  degrees of freedom; Chi2 , chi-squared statistic; p ,  p value; I2 ,  I-squared heterogeneity statistic; Z , Z statistic.

Right ventricular global longitudinal strain in patients with cirrhosis versus patients without cirrhosis

We evaluated six studies [14,15,47,53,55,59] reporting RV-GLS in patients with and without cirrhosis. The pooled analysis of RV-GLS, evaluated using 2D-STE in all studies, revealed a significantly lower absolute RV-GLS based on the mean difference in patients with cirrhosis (MD:-1.95; 95%CI,-3.86 to -0.05; p = 0.04) (Fig 3). REM was considered because there was significant statistical heterogeneity between the studies (I2 = 90%,p<0.00001).

Fig 3

Mean difference of right ventricular global longitudinal strain in patientw with cirrhosis versus patients without cirrhosis evaluated using the random effect model.

SD , standard deviation; IV ,  inverse variance; CI ,  confidence interval; df ,  degrees of freedom; Chi2 , chi-squared statistic; p ,  p-value; I2 ,  I-squared heterogeneity statistic; Z , Z statistic.

We conducted sensitivity analysis to evaluate the stability of our findings. When omitting each study individually from the meta-analysis, the pooled results changed several times (S9 Table). The successive omission of three studies [14,47,55] revealed insignificance in the pooled meta-analysis. Meanwhile, omission of the other three studies [15,53,59] resulted in a significantly lower absolute RV-GLS in favor of the group with cirrhosis. We found that omission of the study by Koç et al. [53] resulted in the largest Q and Tau2 value change. The omission resulted in a lower absolute RV-GLS, favoring the group with cirrhosis (MD:-1.30; 95%CI,-2.43 to -0.18; p = 0.02; I2 = 57%, p = 0.05) (S3 Fig). Further sensitivity analysis was performed, and the study by Rimbaş et al. [15] showed the largest change in Q and Tau2 values (S10 Table). The final result showing no heterogeneity successfuly gives stability in the next sensitivity analysis as shown in S11 Table and resulted in a significantly lower absolute RV-GLS favoring the group with cirrhosisp (MD:-1.90; 95%CI,-2.62 to -1.18; p<0.0001; I2 = 0%, p = 0.76) (S4 Fig). Moreover, to evaluate whether cardiovascular comorbidities could contribute to a decreased RV-GLS, we excluded studies that included patients with hypertension, diabetes mellitus, and left ventricular hypertrophy. Further exclusion of two studies [14,59] with patients with hypertension and diabetes mellitus showed similar results (MD:-2.09; 95%CI,-3.05 to -1.13, p<0.0001; I2 = 0%, p = 0.78). Among these studies, no study had evaluated RV-GLS in patients with left ventricular hypertrophy. Other comorbidities, such as coronary artery disease and valvular heart disease, were excluded in all studies; therefore, these variables did not influence the current results. The effect of diastolic dysfunction on RV-GLS results could not be evaluated because of the limited data available in the remaining study. The effect of ongoing medical therapy (diuretics, beta-blockers, ACE-I/ARB, CCB) on RV-GLS results was evaluated by excluding a study by Chen et al. [14], which still showed a significant reduction in RV-GLS in patients with cirrhosis (MD:-1.97; 95%CI,-2.90 to -1.04, p<0.0001; I2 = 0%, p = 0.57).

Subgroup analysis

Subgroup analysis based on study design showed that when studies were grouped into cross-sectional, case-control, or cohort designs, no significant difference in LV-GLS was found between groups of cross-sectional and cohort studies (S5 Fig). The pooled analysis from all cross-sectional studies did not show a significant difference of LV-GLS between groups (MD:0.16; 95%CI,-2.18–2.50; p = 0.89; I2 = 95%). Pooled analysis of all prospective cohort studies showed similar results (MD:-1.26; 95%CI -3.97–1.46; p = 0.36; I2 = 97%). Meanwhile, a significantly lower LV-GLS in the group with cirrhosis was observed in the pooled analysis of all case-control studies (MD:-3.00; 95%CI,-4.88 to -1.12; p = 0.002; I2 = 90%).

Sensitivity analysis conducted on each study design showed that LV-GLS was significantly lower in the group with cirrhosis (MD:-1.78; 95%CI,-2.50 to -1.07; p<0.0001; I2 = 70%, p = 0.0003) (Fig 5). No significant differences were detected between subgroups (p = 0.18,I2 = 42.5%). Subgroup analysis by study design for RV-GLS was not performed because of the limited number of studies available.

Fig 5

Mean difference of left ventricular global longitudinal strain in patients with cirrhosis versus patients without cirrhosis after subgroup analysis according to study design and consecutive omission of studies by Zamirian, et al.; von Köckritz, et al. and Rimbaş, et al. (cross sectional); consecutive omission of studies by Al-Hwary, et al.; Koç, et al.; and Anish, et al. (case control); and consecutive omission of studies by Altekin, et al.; and Kim, et al (cohort).

SD , standard deviation; IV ,  inverse variance; CI ,  confidence interval; df ,  degrees of freedom; Chi2 , chi-squared statistic; p ,  p value; I2 ,  I-squared heterogeneity statistic; Z , Z statistic.

We did not perform an analysis based on age because all included studies presented age as aggregate information that was reported using mean or median; thus, it could introduce an ecological bias [33]. Subgroup analysis based on sex was not able to be performed because no study reported GLS results separately between male and female patients with cirrhosis.

We did not perform a meta-analysis based on the individual Child-Pugh classification or MELD scores due to the limited number of studies [45,51-53]. Subgroup analysis by dividing cirrhotic patients into compensated and decompensated groups was not possible due to limited studies. There were only two studies [46,50] in which all included patients were classified as having compensated cirrhosis,and three studies [11,48,55] in which all patients included were classified as having decompensated cirrhosis. Comparing GLS changes between studies that included a majority (> 50%) of compensated patients [10,15,46,50,54,59] with studies that included a majority (>50%) of decompensated patients [11,14,36,45,48,49,52,53,55,57,58] may also introduce ecological bias.

Subgroup analysis based on etiology was not performed because of the limited number of studies [15,36,48,53]. There was only one study [47] in which the patient inclusion criteria were solely due to alcoholic etiology, one study [56] was solely due to hepatitis B viral infection, and three studies [46,58,59] solely due to hepatitis C viral infection. Based on individual results from our included studies comparing GLS between various etiologies, two studies reported no significant differences in absolute LV-GLS [48,53] or RV-GLS [53] between different cirrhosis etiologies. However a study by Huang et al. [36] showed that alcoholic etiology had a significantly lower absolute LV-GLS than etiology due to hepatitis B and C infection (20.6±2.3 vs 22.4±2.5 vs 21.9±1.6, respectively; p = 0.034).

Other parameters

We found no significant difference in tricuspid annular plane systolic excursion (TAPSE) between patients with cirrhosis and patients without cirrhosis based on all available data, which was reported by six studies (MD:0.30; 95%CI,-0.25–0.85, p = 0.29; I2 = 0%,p = 0.59) (S6 Fig) [14,15,46,47,49,59]. Similarly, all four studies [14,15,47,59] reported that right ventricular fractional area change (RVFAC) between patients with and without cirrhosis showed no significant difference (MD:-0.56; 95%CI,-1.85–0.73, p = 0.40; I2 = 0%, p = 0.44) (S7 Fig).

Publication bias

A funnel plot (S8 Fig) was generated from the subgroup analysis of studies evaluating LV-GLS. Egger’s regression test revealed that the Z value was 0.490 (p = 0.624), while the rank correlation test indicated a Kendall’s tau of 0.164 (p = 0.542). The results of the trim-and-fill analysis are shown in S9 Fig. The adjusted pooled mean difference with an addition of two studies to make the funnel plot balance was -2.16 (95%CI:-3.00 to -1.32). Overall, a slight publication bias may exist.

Fail-safe N

Based on all the meta-analyses performed, Rosenthal’s fail-safe number showed robust results. Meta-analysis of LV-GLS (Fig 2) and RV-GLS (Fig 3) from all studies showed a fail-safe number of 575 and 123 unpublished studies with null findings, respectively, until a non-significant effect size was obtained. Meta-analysis of 12 studies to reach I2 less than 75% (Fig 4) showed a fail-safe number of 336. The sensitivity analysis conducted on each study design (Fig 5) resulted in a fail-safe number of 317.

Meta-regression

Meta-regression analysis performed to explain variations in the reduction of LV-GLS in the group with cirrhosis revealed that the MELD score and proportion of decompensated cirrhosis covariates were significant and explained R2 = 28% and R2 = 19% of the heterogeneity in LV-GLS reduction, respectively. Figs 6 and 7 show the covariate effect graphs and R2 calculations, respectively. Meta-regression analysis showed that study design, proportion of male subjects, age, proportion of alcoholic etiology, proportion of viral etiology, baseline LVEF, and NOS score had no influence on LV-GLS reduction in the goup with cirrhosis. All data results of the meta-regression are shown in S2–S20 Tables.

Fig 6

Meta regression result for MELD score covariate significantly influencing GLS reduction in cirrhotic group with R2 = 28%.

Fig 7

Meta regression result for decompensated cirrhosis proportion covariate significantly influencing GLS reduction in cirrhotic group with R2 = 19%.

Discussion

The group of patients with cirrhosis had a 1.66% lower LV-GLS after sensitivity analysis was performed by excluding seven studies [11,36,48,50,51,54,57]. Studies by Altekin et al. [54] and Kim et al. [57] contributed the most to the heterogeneity, which may be related to the low methodological quality of the studies. Furthermore, Kim et al. [57] showed a significant age difference between groups; thus, differences in methodology may have affected the outcomes. Meanwhile, a 1.90% lower RV-GLS was found in the group of patients with cirrhosis after two studies were omitted to reach 0% heterogeneity. Heterogeneity sources are from Koç et al. [53] and Rimbaş et al. [15], which might be due to the different study designs and methodological quality.

We realized that the studies included in our analysis had different age ranges, sex proportions, etiology and severity of cirrhosis, and study designs. These variables may have affected the outcomes of our meta-analysis. At present, only subgroup analysis based on study design could be conducted due to the limited number of studies available for subgroup analysis. According to Harrer et al. [33], at least 10 studies should be available for a powerful analysis to show that the subgroups are equivalent, due to the dependence of subgroup analysis on statistical power. After conducting a sensitivity analysis by omitting several studies with low methodological quality, which contributed the most to heterogeneity, the meta-analysis showed that pooled results based on each study design were consistent with the lower LV-GLS in the group of patients with cirrhosis. We reduced heterogeneity by 25% while preserving a meaningful outcome, with a lower absolute LV-GLS of 1.78% in the group with cirrhosis. This indicates that methodological quality and heterogeneity may have affected the results.

We performed meta-regression to investigate the true cause of heterogeneity influencing high I2 values. The MELD score and proportion of decompensated cirrhosis were significantly related to GLS reduction in the group with cirrhosis. These factors may have contributed to the high heterogeneity of the meta-analysis results. We found no significant influence on the results from the study design, sex, age, etiology, baseline LVEF, or NOS score. The finding that cirrhosis severity influences GLS is supported by a previous study [39]. While a lower GLS indicated subclinical systolic dysfunction, a higher GLS was linked to more severe liver disease [39]. Additionally, although some of our meta-analyses showed borderline significant results, our fail-safe number analysis confirmed that all results are robust. This is because the number of studies that are needed to change the result to insignificant exceeded the minimum limit, as proposed by Rosenthal et al [30].

Our study has some limitations. First, due to the limited number of studies, we were unable to provide a subgroup analysis based on age, sex, cirrhosis etiology, and severity. Second, our meta-analysis, which pooled directly from 19 studies for LV-GLS and six studies for RV-GLS, showed considerable statistical heterogeneity. The source of heterogeneity was more likely due to the differences found between studies and not merely due to sampling errors. Major clinical heterogeneity was due to cirrhosis severity based on the meta-regression. Thus, further studies should explore GLS according to the cirrhosis severity.

Conclusion

Our meta-analysis showed that patients with cirrhosis had a lower GLS than controls. However, cirrhosis severity was significantly related to GLS reduction. Future studies should provide further analysis to elucidate a significant difference in GLS in cirrhotic patients according to cirrhosis severity.

(DOCX)

Click here for additional data file.

The mean difference in left ventricular global longitudinal strain from patients with and without cirrhosis after omission of the study by Altekin et al.

(DOCX)

Click here for additional data file.

The mean difference in left ventricular global longitudinal strain between patients with and withour cirrhosis after consecutive omission of studies by Altekin et al. and Kim et al.

(DOCX)

Click here for additional data file.

The mean difference in right ventricular global longitudinal strain from patients with and withour cirrhosis after omission of study by Koç DÖ, et al.

(DOCX)

Click here for additional data file.

The mean difference in right ventricular global longitudinal strain from patients with and without cirrhosis after consecutive omission of studies by Koç DÖ, et al and Rimbaş, et al.

(DOCX)

Click here for additional data file.

The mean difference in left ventricular global longitudinal strain from patients with and withour cirrhosis after subgroup analysis according to study design.

(DOCX)

Click here for additional data file.

The mean difference in tricuspid annular plane systolic excursion between patients with and without cirrhosis.

(DOCX)

Click here for additional data file.

The mean difference in right ventricular fractional area change between patients with and withour cirrhosis.

(DOCX)

Click here for additional data file.

Funnel plot for publication bias on studies evaluating left ventricular global longitudinal strain based on subgroup analysis according to study design in Fig 5.

(DOCX)

Click here for additional data file.

Trim-and-fill analysis for the funnel plot in Fig 5.

GEN generated the mean difference.

(DOCX)

Click here for additional data file.

Meta-regression result for study design covariate have no influence on GLS reduction in group with cirrhosis with R2 = 0%.

(DOCX)

Click here for additional data file.

Meta-regression result for proportion of male subjects covariate have no influence on GLS Reduction in group with cirrhosis with R2 = 0%.

(DOCX)

Click here for additional data file.

Meta-regression result for age covariate have no influence on GLS reduction in group with cirrhosis with R2 = 3%.

(DOCX)

Click here for additional data file.

Meta-regression result for proportion of alcoholic etiology covariate have no influence on GLS reduction in group with cirrhosis with R2 = 0%.

(DOCX)

Click here for additional data file.

Meta-regression result for proportion of viral etiology covariate have no influence on GLS reduction in group with cirrhosis with R2 = 0%.

(DOCX)

Click here for additional data file.

Meta-regression result for baseline left ventricular ejection fraction in group with cirrhosis have no influence on GLS reduction in group with cirrhosis with R2 = 0%.

(DOCX)

Click here for additional data file.

Meta-regression result for Newcastle Ottawa Scale Score have no influence on GLS reduction in group with cirrhosis with R2 = 0%.

(DOCX)

Click here for additional data file.

Search strategy completed using MeSH terms and [all fields].

(DOCX)

Click here for additional data file.

Complete characteristic of included study with left and right ventricular global longitudinal strain results.

(DOCX)

Click here for additional data file.

Newcastle Ottawa Scale for cross-sectional studies.

(DOCX)

Click here for additional data file.

Newcastle Ottawa Scale for case-control studies.

(DOCX)

Click here for additional data file.

Newcastle Ottawa Scale for cohort studies.

(DOCX)

Click here for additional data file.

Sensitivity analysis for mean difference of left ventricular global longitudinal strain from cirrhotic versus non-cirrhotic patients.

(DOCX)

Click here for additional data file.

Sensitivity analysis for mean difference of left ventricular global longitudinal strain from cirrhotic versus non-cirrhotic patients after omission of the study by Altekin et al.

(DOCX)

Click here for additional data file.

Sensitivity analysis for mean difference of left ventricular global longitudinal strain from cirrhotic versus non-cirrhotic patients after omission of studies by Altekin et al. and Kim et al.

(DOCX)

Click here for additional data file.

Sensitivity analysis for mean difference of right ventricular global longitudinal strain from cirrhotic versus non-cirrhotic patients.

(DOCX)

Click here for additional data file.

Sensitivity analysis for mean difference of right ventricular global longitudinal strain from cirrhotic versus non-cirrhotic patients after the mission of study by Koç et al.

(DOCX)

Click here for additional data file.

Sensitivity analysis for mean difference of right ventricular global longitudinal strain from cirrhotic versus non-cirrhotic patients after mission of study by Koç et al.. and Rimbaş et al.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for study design covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for proportion of male subjects covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for age covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for MELD score covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for proportion of decompensated cirrhosis covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for proportion of alcholic etiology covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for proportion of viral etiology covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for baseline left ventricular ejection fraction in cirrhotic group covariate.

(DOCX)

Click here for additional data file.

Meta-regression results and R2 for Newcastle Ottawa Scale score covariate.

(DOCX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

8 Feb 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Vincenzo Lionetti, M.D., PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf  and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Dear Authors,

Thank you for the opportunity to read your work.

After read it with attention the results seems to me sound.

I'm not totally able to assess the statistical analysis, but it appear to me to be right.

Please spent time to siimplify the text that is now too technical.

I will read your revision version with pleasure.

Best Regards

Reviewer #2: The Authors performed a detailed metanalysis showing a significant reduction of LV-GLS and RV-GLS in cirrhotic patients. Overall, they performed an extensive search and analysis from as many as 5347 studies, yielding 20 studies (1738 pts; 2 studies with CMR, 18 with echo). Please find some comments:

-All patients presented by definition a LVEF >50%, but it is unknown how many patients presented cardiovascular comorbidities (hypertension ,diabetes, valvular heart disease, coronary artery disease, LV hypertrophy....) that might contribute to a decreased LV or RV GLS

-It is not clear which was the relationship between LVGLS and LV ejection fraction/MAPSE, or between RVGLS and TAPSE/FAC; did you analyse this point? Was there a difference between patients with bordeline LVEF and normal/high LVEF?

-Besides an analysis on compensated/decompensated liver cirrhosis, were there differences according to ongoing medical therapy (diuretics? beta-blockers? ACE/ARB? MRA?) or to the presence of heart failure?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Luigi Vetrugno

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

16 Mar 2022

Response from Academic Editor:

ACADEMIC EDITOR: All issues raised by expert reviewers are required. Moreover, the authors should revise statistical approach in order to better support the conclusions.

Response from Author:

We have revise our statistical approach. The reporting of the trial sequential analysis has been improved by defining the mean difference a priori, before the trial sequential analysis run while also correcting the heterogeneity to reach a low heterogeneity (at least 25%). With this approach we report a more robust result. This was further described in the results section (Trial Sequential Analysis) and indicated in Figure S10 and Figure S11. This means that all result from meta analysis have shown a conclusive evidence with a low risk of a false positive result.

We calculated the information size required based on α = 5% significance level, and β = 20% (80% power) with an anticipated LV-GLS change of -1.43% and RV-GLS change of -1.95% in favor of cirrhotic groups. Heterogeneity was corrected to reach I2 of at least 25% (low heterogeneity). [PAGE 4]

TSA result for LV-GLS and RV-GLS are shown in S10 Fig and S11 Fig, respectively. Both TSA showed that these studies cross trial monitoring boundary and confirm a lower absolute GLS in favor of cirrhotic group. This indicate that result of GLS in both ventricle have low risk of false positive results. After conducting TSA with sensitivity analysis that has been done previously in Fig 4 for LV-GLS and Figure S4 for RV-GLS, a further TSA is then generated for LV-GLS (S12 Fig) and RV-GLS (S13 Fig). From these TSA, the result is also conclusive and have low risk of false positive results. [PAGE 12]

The conclusion have been better improved in accordance with the improved results by using a revised statistical approach

Despite the clinical and methodological heterogeneity, TSA revealed that current meta analysis have shown firm evidences that cirrhotic patients have a lower GLS when compared with controls. Future studies should provide a further analysis to elucidate whether there will be a significant difference of GLS in cirrhotic patients with different age, sex, cirrhosis severity and etiologies. [PAGE 16]

Response from Reviewer:

Reviewer #1: Dear Authors, Thank you for the opportunity to read your work. After read it with attention the results seems to me sound. I'm not totally able to assess the statistical analysis, but it appear to me to be right. Please spent time to siimplify the text that is now too technical. I will read your revision version with pleasure.

Response from Authors:

We have simplify the text particularly the discussion part in order to be more readable. Several grammatical errors have also been corrected. [PAGE 2-16]

Response from Reviewer:

Reviewer #2: The Authors performed a detailed metanalysis showing a significant reduction of LV-GLS and RV-GLS in cirrhotic patients. Overall, they performed an extensive search and analysis from as many as 5347 studies, yielding 20 studies (1738 pts; 2 studies with CMR, 18 with echo). Please find some comments:

-All patients presented by definition a LVEF >50%, but it is unknown how many patients presented cardiovascular comorbidities (hypertension ,diabetes, valvular heart disease, coronary artery disease, LV hypertrophy....) that might contribute to a decreased LV or RV GLS

Response from Author:

We have added additional sensitivity analysis related to cardiovascular comorbidities on its influence toward LV-GLS and RV-GLS result.

Moreover, in order to evaluate whether cardiovascular comorbidities could have contribution to a decreased LV-GLS, we excluded studies that include patients with hypertension, diabetes mellitus, and left ventricular hypertrophy. Further exclusion of three studies[14,44,57] with hypertension and diabetes mellitus patients showed similar result (MD:-1.55;95%CI,-2.37 – -0.73,p=0.0002; I2=73%,p=0.0003) while exclusion of a study[43] with left ventricular hypertrophy patients also showed simillar results (MD:-1.59;95%CI,-2.34 – -0.84,p<0.00001; I2=70%,p=0.0002). Other comorbidities such as coronary artery disease and valvular heart disease were excluded by all studies and therefore these variable did not influence current result. [PAGE 12]

Moreover, in order to evaluate whether cardiovascular comorbidities could have contribution to a decreased RV-GLS, we excluded studies that include patients with hypertension, diabetes mellitus, and left ventricular hypertrophy. Further exclusion of two studies[14,57] with hypertension and diabetes mellitus patients showed similar result (MD:-2.09;95%CI,-3.05 – -1.13,p<0.0001; I2=0%,p=0.78). Among these studies, there were no study evaluating RV-GLS which include patients with left ventricular hypertrophy. Other comorbidities such as coronary artery disease and valvular heart disease were excluded by all studies and therefore these variable did not influence current result. [PAGE 13]

Response from Reviewer:

-It is not clear which was the relationship between LVGLS and LV ejection fraction/MAPSE, or between RVGLS and TAPSE/FAC; did you analyse this point? Was there a difference between patients with bordeline LVEF and normal/high LVEF?

Response from Author:

We have added additional analysis comparing TAPSE and FAC between groups. Additional figure are added in suplementary files. However, there was no study reporting MAPSE and thus evaluation are limited.

Other Parameters

We found no significant difference of tricuspid annular plane systolic excursion (TAPSE) between cirrhotic versus non-cirrhotic patients based on all available data which was reported by 6 studies (MD:0.30;95%CI,-0.25 – 0.85, p=0.29; I2=0%,p=0.59) (S6 Fig).[14,15,44,45,47,57] Similarly, all 4 studies[14,15,45,57] reporting right ventricular fractional area change (RVFAC) between cirrhotic and non-cirrhotic patients showed no significant difference (MD:-0.56;95%CI,-1.85 – 0.73, p=0.40; I2=0%,p=0.44) (S7 Fig). [PAGE 14]

Response from Reviewer:

-Besides an analysis on compensated/decompensated liver cirrhosis, were there differences according to ongoing medical therapy (diuretics? beta-blockers? ACE/ARB? MRA?) or to the presence of heart failure?

Response from Author:

We have added additional sensitivity analysis to evaluate the influence of heart failure (diastolic dysfunction, abnormal NT-proBNP result) and ongoing medical therapy that may affect the LV-GLS result. Moreover, we have update Table 1 by adding information regarding ongoing medical therapy and NT-proBNP. No study reporting the use of MRA in the population included and thus these variable was unable to be evaluated.

Additionaly, exclusion of two studies[15,47] which include patients with diastolic dysfunction revealed a similar result indicating a significant reduction of LV-GLS in cirrhotic patients (MD:-1.85;95%CI,-2.57 – -1.33,p<0.00001; I2=63%,p=0.003). In order to evaluate whether ongoing medical therapy (diuretics, beta-blockers, angiotensin converting enzyme inhibitor – ACE-I/angiotensin receptor blocker – ARB, calcium channel blocker – CCB) could influence LV-GLS, we exclude 3 studies[10,14,43] that report these medications in cirrhotic patients and the result showed that there is still a significant reduction of LV-GLS in cirrhotic patients (MD:-1.69;95%CI,-2.59 – -0.79,p=0.0002; I2=76%,p<0.0001). The effect of diuretic treatment that can be seen when a study[10] was excluded from the analysis also resulting in a similar result (MD:-1.72;95%CI,-2.40 – -1.04,p<0.00001; I2=71%,p=0.0002). Excluding a study[43] to evaluate the influence of beta-blocker treatment on the result also showing similar finding (MD:-1.59;95%CI,-2.34 – -0.84,p<0.0001; I2=70%,p=0.0002). [PAGE 12]

The effect of diastolic dysfunction on RV-GLS results is unable to be evaluated because limited data availble on the remaining study. The effect of ongoing medical therapy (diuretics, beta-blockers, ACE-I/ARB, CCB) on RV-GLS result was evaluated by excluding a study[14] which still showing a significant reduction of RV-GLS in cirrhotic patients (MD:-1.97;95%CI,-2.90 – -1.04,p<0.0001; I2=0%,p=0.57). [PAGE 13]

Submitted filename: REVISION NOTES PLOS ONE.docx

Click here for additional data file.

21 Mar 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Vincenzo Lionetti, M.D., PhD

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

22 Mar 2022

REVISION#1

Academic Editor

Response from Academic Editor:

All issues raised by expert reviewers are required. Moreover, the authors should revise statistical approach in order to better support the conclusions.

Response from Author:

We have revise our statistical approach. The reporting of the trial sequential analysis has been improved by defining the mean difference a priori, before the trial sequential analysis run while also correcting the heterogeneity to reach a low heterogeneity (at least 25%). With this approach we report a more robust result. This was further described in the results section (Trial Sequential Analysis) and indicated in Figure S10 and Figure S11. This means that all result from meta analysis have shown a conclusive evidence with a low risk of a false positive result.

We calculated the information size required based on α = 5% significance level, and β = 20% (80% power) with an anticipated LV-GLS change of -1.43% and RV-GLS change of -1.95% in favor of cirrhotic groups. Heterogeneity was corrected to reach I2 of at least 25% (low heterogeneity). [PAGE 4]

TSA result for LV-GLS and RV-GLS are shown in S10 Fig and S11 Fig, respectively. Both TSA showed that these studies cross trial monitoring boundary and confirm a lower absolute GLS in favor of cirrhotic group. This indicate that result of GLS in both ventricle have low risk of false positive results. After conducting TSA with sensitivity analysis that has been done previously in Fig 4 for LV-GLS and Figure S4 for RV-GLS, a further TSA is then generated for LV-GLS (S12 Fig) and RV-GLS (S13 Fig). From these TSA, the result is also conclusive and have low risk of false positive results. [PAGE 12]

The conclusion have been better improved in accordance with the improved results by using a revised statistical approach

Despite the clinical and methodological heterogeneity, TSA revealed that current meta analysis have shown firm evidences that cirrhotic patients have a lower GLS when compared with controls. Future studies should provide a further analysis to elucidate whether there will be a significant difference of GLS in cirrhotic patients with different age, sex, cirrhosis severity and etiologies. [PAGE 16]

Journal Requirements

Response from Journal:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Response from Author:

We have followed all guidance accordingly

Response from Journal:

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

Response from Author:

We have update Data Availability information. We have attached all of our raw data (File Name: RAW DATA_PLOS ONE). No ethical or legal restriction in sharing our data.

Response from Journal:

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

Response from Author:

We have added captions for our Supporting Information files at the end of manuscript.

Reviewers' Comments

Reviewer's Responses to Questions

Comments to the Author:

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Response from Author:

We thank all reviewers that have provided responses indicating that our manuscript is technically sound and that the data have support the conclusion

Comments to the Author:

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: I Don't Know

Response from Author:

In reporting our systematic review, we followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 statement guideline. (Reference Number 18)

We comply with Conducting Systematic Reviews and Meta-Analyses of Observational Studies of Etiology (COSMOS-E) guidance to undertake this systematic review and meta-analysis. (Reference Number 19)

Report of statistical heterogeneity and publication bias was in accordance with the Cochrane Handbook for Systematic Reviews of Interventions version 6.1. (Reference Number 22)

Sensitivity analysis was in accordance with a guideline from “Doing a Meta-Analysis with R: A Hands on Guide” (Reference Number 30)

Trial Sequential Analysis was performed in accordance with User manual for trial sequential analysis (TSA). (Reference Number 38)

Methodological quality evaluation was in accordance with The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. (Reference Number 40)

All statistical analysis was performed using Review Manager 5.3.5 software (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration) and JASP Version 0.14.1

We believe by following all of these guidelines, our statistical analysis has been performed appropriately and rigorously. Moreover, we have attached all of our raw data (File Name: RAW DATA_PLOS ONE) in which all of these data could be utilized to replicate the reported study findings in their entirety.

Comments to the Author:

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Response from Author:

We thank all reviewers that have provided responses indicating that the authors have made all data underlying the findings in manuscript fully available

Comments to the Author:

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Response from Author:

We thank all reviewers that have provided responses indicating that the manuscript is presented in an intelligible fashion and written in standard English.

Response from Reviewer:

Reviewer #1: Dear Authors, Thank you for the opportunity to read your work. After read it with attention the results seems to me sound. I'm not totally able to assess the statistical analysis, but it appear to me to be right. Please spent time to siimplify the text that is now too technical. I will read your revision version with pleasure.

Response from Authors:

We have simplify the text particularly the discussion part in order to be more readable. Several grammatical errors have also been corrected. [PAGE 2-16]

Response from Reviewer:

Reviewer #2: The Authors performed a detailed metanalysis showing a significant reduction of LV-GLS and RV-GLS in cirrhotic patients. Overall, they performed an extensive search and analysis from as many as 5347 studies, yielding 20 studies (1738 pts; 2 studies with CMR, 18 with echo). Please find some comments:

-All patients presented by definition a LVEF >50%, but it is unknown how many patients presented cardiovascular comorbidities (hypertension ,diabetes, valvular heart disease, coronary artery disease, LV hypertrophy....) that might contribute to a decreased LV or RV GLS

Response from Author:

We have added additional sensitivity analysis related to cardiovascular comorbidities on its influence toward LV-GLS and RV-GLS result.

Moreover, in order to evaluate whether cardiovascular comorbidities could have contribution to a decreased LV-GLS, we excluded studies that include patients with hypertension, diabetes mellitus, and left ventricular hypertrophy. Further exclusion of three studies[14,44,57] with hypertension and diabetes mellitus patients showed similar result (MD:-1.55;95%CI,-2.37 – -0.73,p=0.0002; I2=73%,p=0.0003) while exclusion of a study[43] with left ventricular hypertrophy patients also showed simillar results (MD:-1.59;95%CI,-2.34 – -0.84,p<0.00001; I2=70%,p=0.0002). Other comorbidities such as coronary artery disease and valvular heart disease were excluded by all studies and therefore these variable did not influence current result. [PAGE 12]

Moreover, in order to evaluate whether cardiovascular comorbidities could have contribution to a decreased RV-GLS, we excluded studies that include patients with hypertension, diabetes mellitus, and left ventricular hypertrophy. Further exclusion of two studies[14,57] with hypertension and diabetes mellitus patients showed similar result (MD:-2.09;95%CI,-3.05 – -1.13,p<0.0001; I2=0%,p=0.78). Among these studies, there were no study evaluating RV-GLS which include patients with left ventricular hypertrophy. Other comorbidities such as coronary artery disease and valvular heart disease were excluded by all studies and therefore these variable did not influence current result. [PAGE 13]

Response from Reviewer:

-It is not clear which was the relationship between LVGLS and LV ejection fraction/MAPSE, or between RVGLS and TAPSE/FAC; did you analyse this point? Was there a difference between patients with bordeline LVEF and normal/high LVEF?

Response from Author:

We have added additional analysis comparing TAPSE and FAC between groups. Additional figure are added in suplementary files. However, there was no study reporting MAPSE and thus evaluation are limited.

Other Parameters

We found no significant difference of tricuspid annular plane systolic excursion (TAPSE) between cirrhotic versus non-cirrhotic patients based on all available data which was reported by 6 studies (MD:0.30;95%CI,-0.25 – 0.85, p=0.29; I2=0%,p=0.59) (S6 Fig).[14,15,44,45,47,57] Similarly, all 4 studies[14,15,45,57] reporting right ventricular fractional area change (RVFAC) between cirrhotic and non-cirrhotic patients showed no significant difference (MD:-0.56;95%CI,-1.85 – 0.73, p=0.40; I2=0%,p=0.44) (S7 Fig). [PAGE 14]

Response from Reviewer:

-Besides an analysis on compensated/decompensated liver cirrhosis, were there differences according to ongoing medical therapy (diuretics? beta-blockers? ACE/ARB? MRA?) or to the presence of heart failure?

Response from Author:

We have added additional sensitivity analysis to evaluate the influence of heart failure (diastolic dysfunction, abnormal NT-proBNP result) and ongoing medical therapy that may affect the LV-GLS result. Moreover, we have update Table 1 by adding information regarding ongoing medical therapy and NT-proBNP. No study reporting the use of MRA in the population included and thus these variable was unable to be evaluated.

Additionaly, exclusion of two studies[15,47] which include patients with diastolic dysfunction revealed a similar result indicating a significant reduction of LV-GLS in cirrhotic patients (MD:-1.85;95%CI,-2.57 – -1.33,p<0.00001; I2=63%,p=0.003). In order to evaluate whether ongoing medical therapy (diuretics, beta-blockers, angiotensin converting enzyme inhibitor – ACE-I/angiotensin receptor blocker – ARB, calcium channel blocker – CCB) could influence LV-GLS, we exclude 3 studies[10,14,43] that report these medications in cirrhotic patients and the result showed that there is still a significant reduction of LV-GLS in cirrhotic patients (MD:-1.69;95%CI,-2.59 – -0.79,p=0.0002; I2=76%,p<0.0001). The effect of diuretic treatment that can be seen when a study[10] was excluded from the analysis also resulting in a similar result (MD:-1.72;95%CI,-2.40 – -1.04,p<0.00001; I2=71%,p=0.0002). Excluding a study[43] to evaluate the influence of beta-blocker treatment on the result also showing similar finding (MD:-1.59;95%CI,-2.34 – -0.84,p<0.0001; I2=70%,p=0.0002). [PAGE 12]

The effect of diastolic dysfunction on RV-GLS results is unable to be evaluated because limited data availble on the remaining study. The effect of ongoing medical therapy (diuretics, beta-blockers, ACE-I/ARB, CCB) on RV-GLS result was evaluated by excluding a study[14] which still showing a significant reduction of RV-GLS in cirrhotic patients (MD:-1.97;95%CI,-2.90 – -1.04,p<0.0001; I2=0%,p=0.57). [PAGE 13]

Comments to the Author:

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Luigi Vetrugno

Reviewer #2: No

Response from Author:

We thank all the reviewers for providing comments to our manuscript. We have provided feedback to all of the reviewers comments point by point.

REVISION #2

Academic Editor

Response from Academic Editor:

All issues raised by expert reviewers are required.

Response from Authors:

We have revised our attachment file in “Response to Reviewers” section in order to provide a more complete response to all of the statements given in the previous Revision round. We can not find any additional issues raised by reviewers in this second revision within the letter sent on March 21 2022 01:53PM. In this second round revision, we also did not find any action link (“View Attachment”) as indicated by the journal if reviewer comments were submitted as an attachment file. Please let us know if there is an error that make us unable to find additional issues from reviewers in the letter provided on March 21 2022 01:53PM.

Response from Journal:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Response from Author:

We have used PACE to ensure that figures meet PLOS requirments.

Submitted filename: Response to Reviewer_Revision 1-2_PLOS ONE.docx

Click here for additional data file.

11 Apr 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Vincenzo Lionetti, M.D., PhD

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have address the issue raise by the reviewers. Now the manuscript appear clear and the statistical part is revised.

Best Regards

Reviewer #2: (No Response)

Reviewer #3: The systematic review appears to be rather thorough. The meta-analysis is carried out routinely as one would expect. The publication bias and sensitivity analysis are performed as one would expect. The results appear to follow from the analyses.

The confusion comes in the inclusion of the studies and their exact description from a statistical perspective. Table 1 is not much help. Although described in the text, there are 12 case control and cross sectional studies and 8 prospective studies. The prospective cohort are not well described. Exactly how many studies had a direct randomized comparison? In order to make valid statistical comparisons overall, these should be looked at separately.

The addition of the TSA adds further confusion. It’s exact purpose is not clearly stated. How does it add statistically to the strength of the meta-analytic results?

As seen in Figures 2 to 4. The heterogeneity is rather high especially in Figures 2 and 3. The investigators do not seem to have investigated the true causes of heterogeneity and explain such a high value of the I-square statistic. Also, the overall p-value of the p=0.04 is borderline significant. The overall fail safe number of publications to give assurance to these values should have been calculated.

The authors overwhelm the reader with many analyses and sub analyses and one can lose their focus with such a confusing statistical presentation.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Luigi Vetrugno

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

22 May 2022

REVISION #3

Academic Editor

Response from Academic Editor:

Statistical revision was performed by expert in addition to selected reviewers in order to improve the quality of the manuscript. Major issues need to be adequately solved by the authors as reported by the third reviewer. These issues are required.

Response from Author:

We have revise our statistical approach:

1. We have done meta regression in order to investigate the true causes of heterogeneity and provide explaination of factor that associated with high I-square value. (PAGE 4, LINE 129; PAGE 16, LINE 370; PAGE 17, LINE 397)

2. We have removed TSA to reduce further confusion due to many statistical presentation (PAGE 4, LINE 148)

3. We have added meta analysis report based on study design separately (PAGE 14, LINE 318)

4. A fail safe N analysis have been added to give assurance to the borderline significant statistic. Result shows a robust finding. (PAGE 4, LINE 124; PAGE 16, LINE 363; PAGE 17, LINE 402)

Reviewers' Comments

Reviewer's Responses to Questions

Comments to the Author:

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

Reviewer #3: (No Response)

Response from Author:

We thank all reviewers that have provided responses indicating that our manuscript have addressed all comments

Comments to the Author:

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Partly

Response from Author:

We thank all reviewers that have provided responses indicating that our manuscript is technically sound and the data support the conclusions. Additionaly, we have revised our manuscript following Reviewer #3 comments.

Comments to the Author:

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

Response from Author:

We thank all reviewers that have provided responses indicating that the statistical analysis has been performed appropriately and rigorously. Additionaly, we have revised our manuscript following Reviewer #3 comments.

Comments to the Author:

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

Response from Author:

We thank all reviewers that have provided responses indicating that the authors have made all data underlying the findings in manuscript fully available

Comments to the Author:

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

Response from Author:

Following Reviewer #3 response, we have sent our manuscript to Editage English Editing Service to improve our manuscript writing and comply with the standard for English scientific writing. We have uploaded the certificate of editing with the file name “Certificate_of_editing-DEIDJ_1”.

Response from Reviewer:

6. Review Comments to the Author

Reviewer #1: The authors have address the issue raise by the reviewers. Now the manuscript appear clear and the statistical part is revised. Best Regards.

Response from Author:

We thank Reviewer #1 for all the comments indicating that our manuscript appear clear and the statistical part is revised.

Response from Reviewer:

Reviewer #2: (No Response)

Response from Author:

We thank Reviewer #2 for all the comments in the previous revision.

Response from Reviewer:

Reviewer #3: The systematic review appears to be rather thorough. The meta-analysis is carried out routinely as one would expect. The publication bias and sensitivity analysis are performed as one would expect. The results appear to follow from the analyses.

Response from Author:

We thank Reviewer #3 that have provided responses indicating that our manuscript appears to be rather thorough, the meta analysis was carried out routinely as expected, publication bias and sensitivity analysis performed as expected, and results appear to follow from the analyses.

Response from Reviewer:

Reviewer #3: The confusion comes in the inclusion of the studies and their exact description from a statistical perspective.

Response from Author:

For studies inclusion, the description of all findings have been detailed based on study design. (PAGE 5, LINE 166; PAGE 12, LINE 206). For statistical revision we have performed the following:

1. We have done meta regression in order to investigate the true causes of heterogeneity and provide explaination of factor that associated with high I-square value. (PAGE 4, LINE 129; PAGE 16, LINE 370; PAGE 17, LINE 397)

2. We have removed TSA to reduce further confusion due to many statistical presentation (PAGE 4, LINE 148)

3. We have added meta analysis report based on study design separately (PAGE 14, LINE 318)

4. A fail safe N analysis have been added to give assurance to the borderline significant statistic. Result shows a robust finding. (PAGE 4, LINE 124; PAGE 16, LINE 363; PAGE 17, LINE 402)

Response from Reviewer:

Table 1 is not much help.

Response from Author:

Table 1 have been simplified.

Response from Reviewer:

Although described in the text, there are 12 case control and cross sectional studies and 8 prospective studies.

The description of all findings have been detailed based on study design. (PAGE 5, LINE 166 – LINE 181)

“This systematic review identified... Other comorbidities such as dyslipidemia and pulmonary artery hypertension were reported in one cross-sectional [46] and one case-control study[51], respectively.”

The description of all findings have been detailed based on study design. (PAGE 12, LINE 206 – LINE 216)

“For left ventricular assessment, there were two cross-sectional… included patients with a borderline EF (50-55%).”

Response from Reviewer:

The prospective cohort are not well described.

Response from Author:

We have extensively added description of all prospective cohort studies. (PAGE 5, LINE 182 – LINE 200)

“All five prospective cohorts had different durations… The mean MELD score was not reported.”

Response from Reviewer:

Exactly how many studies had a direct randomized comparison? In order to make valid statistical comparisons overall, these should be looked at separately.

Response from Author:

We have added new statement that our search strategy did not identify any randomized controlled trial. (PAGE 5, LINE 168)

“Our search strategy did not identify any randomized controlled trials.”

Response from Reviewer:

The addition of the TSA adds further confusion. It’s exact purpose is not clearly stated. How does it add statistically to the strength of the meta-analytic results?

Response from Author:

We have removed TSA to reduce further confusion due to many statistical presentation. (PAGE 4, LINE 148)

Response from Reviewer:

As seen in Figures 2 to 4. The heterogeneity is rather high especially in Figures 2 and 3. The investigators do not seem to have investigated the true causes of heterogeneity and explain such a high value of the I-square statistic.

Response from Author:

We have done meta regression in order to investigate the true causes of heterogeneity and provide explaination of factor that associated with high I-square value. (METHODS, PAGE 4, LINE 129 – LINE 135)

“We performed a random-effects (method of moments) meta-regression… baseline LVEF in the group with cirrhosis, and methodological quality of the study.”

We have done meta regression in order to investigate the true causes of heterogeneity and provide explaination of factor that associated with high I-square value. (RESULT, PAGE 16, LINE 370 – LINE 377)

“Meta-regression analysis performed to explain variations in the reduction… All data results of the meta-regression are shown in S12-S20 Table.”

We have done meta regression in order to investigate the true causes of heterogeneity and provide explaination of factor that associated with high I-square value. (DISCUSSION, PAGE 17, LINE 397 – LINE 402)

“We performed meta-regression to investigate the true cause of heterogeneity influencing… a higher GLS was linked to more severe liver disease.”

Response from Reviewer:

Also, the overall p-value of the p=0.04 is borderline significant. The overall fail safe number of publications to give assurance to these values should have been calculated.

Response from Author:

A fail safe N analysis have been added to give assurance to the borderline significant statistic. Result shows a robust finding. (METHODS, PAGE 4, LINE 124 – LINE 128)

“The overall fail-safe number of publications… five times higher than the number of studies included, plus 10”

A fail safe N analysis have been added to give assurance to the borderline significant statistic. Result shows a robust finding. (RESULTS, PAGE 16, LINE 363 – LINE 367)

“Based on all the meta-analyses performed, Rosenthal’s fail-safe number… resulted in a fail-safe number of 317.”

A fail safe N analysis have been added to give assurance to the borderline significant statistic. Result shows a robust finding. (DISCUSSION, PAGE 17, LINE 402 – LINE 405)

“Additionally, although some of our meta-analyses… exceeded the minimum limit, as proposed by Rosenthal et al.”

Response from Reviewer:

The authors overwhelm the reader with many analyses and sub analyses and one can lose their focus with such a confusing statistical presentation.

Response from Author:

We have removed TSA to reduce further confusion due to many statistical presentation (PAGE 4, LINE 148).

Comments to the Author:

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Luigi Vetrugno

Reviewer #2: No

Reviewer #3: No

Response from Author:

We thank all the reviewers for providing comments to our manuscript. We have provided feedback to all of the reviewers comments point by point.

Submitted filename: Response to Reviewer_Revision 1-3_PLOS ONE.docx

Click here for additional data file.

26 May 2022

Evaluation of subclinical ventricular systolic dysfunction assessed using global longitudinal strain in liver cirrhosis: a systematic review, meta-analysis, and meta-regression

PONE-D-21-37272R3

Dear Dr. Ridjab,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Vincenzo Lionetti, M.D., PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

Reviewer #3: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for addressing my comments.

This is R3 and I have not further question to you.

Best Regards

Reviewer #2: (No Response)

Reviewer #3: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Luigi Vetrugno

Reviewer #2: No

Reviewer #3: No

30 May 2022

PONE-D-21-37272R3

Evaluation of subclinical ventricular systolic dysfunction assessed using global longitudinal strain in liver cirrhosis: a systematic review, meta-analysis, and meta-regression

Dear Dr. Ridjab:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Vincenzo Lionetti

Academic Editor

PLOS ONE